Friday April 8 2011 (

Alcohol was thought caused 58,000 cancers in the study group


Alcohol “causes 13,000 cancer cases a year”, The Daily Telegraph has reported. The newspaper says that in the UK drinking is responsible for 2,500 cases of breast cancers, 3,000 bowel cancers and 6,000 cases of cancers of the mouth, throat or windpipe.

The research used data from a large European study which looked at how current and former alcohol consumption related to the development of cancer in more than 350,000 people from eight countries. The researchers extrapolated the results to the general population and estimated that, across Europe, 10% of all cancers in men and 3% of all cancers in women could be attributed to alcohol consumption. There was a stronger association with cancers that are already known to be causally associated with alcohol, such as cancers of the mouth, throat, oesophagus and liver. For these cancers, much of the excess risk was due to drinking above the maximum daily limit, defined in this study as more than 24g of pure alcohol for men (3 units) and more than 12g for women (1.5 units).

In the UK, the current recommended daily limit for men is 3-4 units, while for women it is no more than 2-3 units daily. One unit is equivalent to 8g alcohol, or about half a pint of weak lager.

Where did the story come from?

The study was carried out by researchers from the German Institute of Human Nutrition in Potsdam-Rehbruecke, and other institutions in Europe and the US. It received funding from numerous organisations and was published in the peer-reviewed British Medical Journal.

News coverage has reflected the findings of this well-conducted study.

What kind of research was this?

This was a cohort study which aimed to establish alcohol’s contribution towards the burden of cancer across eight European countries. To do this, researchers used data from the European Prospective Investigation into Cancer and Nutrition (EPIC), a large cohort study set up to examine how the diet and lifestyle of a large European population sample related to their development of cancer over a  follow-up period of nearly nine years.

Alongside their alcohol-related findings based on this cohort, the researchers also used general population-based data on alcohol consumption and cancer incidence to extrapolate the findings to the national populations from which the EPIC participants were drawn.

What did the research involve?

The EPIC study started in 1992 and recruited 520,000 men and women (aged from 37 to 70 years) from the general population of 10 European countries: France, Italy, Spain, Holland, Greece, Germany Denmark, Norway, Sweden and the UK.

Upon entry into the study, the participants completed diet and lifestyle questionnaires. The researchers excluded people with cancer at the start of the study and those with missing questionnaire data on alcohol consumption. This lead the researchers to include 109,118 men and 254,870 women across eight countries in their analysis (data from Norway and Sweden could not be used due to a lack of data on past alcohol consumption.

The validated questionnaires asked participants to estimate their alcohol consumption in the year before recruitment, both in terms of average grams of pure alcohol per day and as the frequency/portion size of beer, wine, spirits, etc. Researchers also asked about past consumption at ages 20, 30, 40 and 50. Based on these two responses – past and current consumption – people were categorised as:

  • never drinkers – no consumption in the past or at recruitment
  • former drinkers – consumption in the past but no consumption at recruitment
  • lifetime drinkers – consumption both in the past and at recruitment

Cancer outcomes for each individual were assessed up to the years 2000-2005, using regional cancer registries, checks of medical records, health insurance records, pathology records and death certificates. The precise methods varied according to the practices used in each country. The mean follow-up time was almost nine years.

Risk associations between cancer and current and former alcohol use were conducted separately for men and women. The researchers made adjustments to account for the influence of numerous potential socioeconomic and lifestyle confounders, including smoking, diet, BMI and education level. The risk figures obtained for the association between alcohol and cancer were then applied to alcohol consumption in each country’s general population (calculated from World Health Organization surveys and per capita consumption data) and cancer incidence data to estimate the total number of cases of cancer per year that could be attributable to alcohol in males and females aged 15 and above.

Researchers used this data to calculate a measure called the ‘population attributable fraction’ for consumption beyond the recommended daily upper limit, which would estimate what proportion of cancer cases were associated with drinking more than 24g of pure alcohol day for men (equivalent to 3 units) and 12g/day alcohol for women (equivalent to 1.5 units). The population attributable fraction indicates what reduction in cancer incidence would be expected if consumption were reduced below this level.

What were the basic results?

There was variation in average alcohol consumption across European countries. By applying the results of the EPIC study to national population data, the study estimated that 10% of all cancers in men in Europe (95% confidence interval 7 to 13%) and 3% of all cancers in women in Europe (1 to 5%) could be attributable to alcohol consumption (both former and current).

The researchers also calculated alcohol attributable fractions relating to specific cancers:

  • Upper digestive tract cancers (e.g. mouth, throat, oesophagus) – 44% for males and 25% for females
  • Liver cancer – 33% for males and 18% for females
  • Colorectal cancer – 17% for males and 4% for females
  • Female breast cancer – 5% of cases.
    UK-specific data were similar to these European averages.

Based on 2008 European cancer data, alcohol consumption above the daily maximum (as defined above) caused 33,037 of 178,578 alcohol-related cancers in men (18.5%) and 17,470 of the 397,043 alcohol-related cancers in women (4.4%).

How did the researchers interpret the results?

The researchers conclude that an “important proportion” of cancers in Western Europe can be attributed to alcohol consumption, especially when consumption is higher than the recommended daily upper limits. They say that their data “supports current political efforts to reduce or to abstain from alcohol consumption to reduce the incidence of cancer”.


This study has assessed the association between alcohol consumption and cancer risk, and it has estimated how the cancer burden could be reduced by lowering consumption to below the daily maximum limits (defined in this study as 24g for men and 12g for women). The study has several strengths, including its large study population drawn from eight European countries and thorough follow-up of participants (less than 2% of the sample in all countries were lost during the follow-up process). It also combined the cohort data with general population data on alcohol consumption and cancer figures to estimate country-relevant data.

There are some limitations that should be acknowledged:

  • The underlying data on alcohol intake was self-reported by participants, and the quality of the consumption data would rely on them accurately estimating their drinking. The study also looked at consumption during past decades, which might be particularly difficult to recall.
  • The study may not have adjusted for all possible confounders (i.e. factors that are linked to both alcohol consumption and cancer outcomes). However, they did adjust for the most obvious ones, which is a strength of this cohort.
  • The researchers say the estimates they calculated in this study were based on an assumption that alcohol is causative in the cancers studied (e.g.cancers of the aerodigestive system and liver). While alcohol may not be conclusively proven as a cause of these cancers there is a great deal of evidence suggesting that this is a major cause.
  • There may be differences between the participants who agreed to participate and those who did not. If this is so, the results may not be generalisable to the populations from which the samples were drawn.
  • The study looked at people who drank beyond recommended daily limits, but did not calculate how increasing levels of consumption related to cancer risk.

The study estimates that, in the European population as a whole, 10% of all cancers in men and 3% of all cancers in women could be attributed to alcohol consumption. Alcohol consumption was already known to be associated with numerous cancers, in particular those of the mouth, throat, oesophagus, liver and bowel, and this study data supports those associations. For those cancers that are believed to be causatively associated with cancer, the study estimates that 32% in men and 5% in women can be attributed to alcohol, and a large proportion of this attributable fraction is due to consumption above the daily maximum.

As the researchers appropriately conclude, there is a “necessity to continue and to increase efforts to reduce alcohol consumption in Europe, both on the individual and the population level”.